Recent discoveries about the molecular organization and physical properties of the mammalian erythrocyte membrane and its associated structural proteins can now be used to explain, and may eventually be used to predict, the shape of the erythrocyte. Such explanations are possible because the relatively few structural proteins of the erythrocyte are regularly distributed over the entire cytoplasmic surface of the cell membrane and because the well-understood topological associations of these proteins seem to be stable in comparison with the time required for the cell to change shape. These simplifications make the erythrocyte the first nonmuscle cell for which it will be possible to extend our knowledge of molecular interactions to the next hierarchical level of organization that deals with shape and shape transformations.